The understanding of the biochemical synthetic pathways in the metabolism of animal or plant cells, including microorganisms such as bacteria, fungi and algae, or mammalian cells, remains very rudimentary, even though the main synthetic pathways are known. To date, the determination of physiological states during growth, development or as a response to environmental stress is essentially limited to the study of individual target molecules such as, for example, RNA and proteins. However, changes in the mRNA or protein level or their activity can frequently not be correlated with changes in the metabolism or indeed with phenotypic functions.
Cellular constituents or metabolites are frequently analyzed directly either by specific enzymatic reactions, immunoassays or on the basis of chromatographic methods, which identify certain substances by their retention times or coelution with reference substances. As described in Katona, J. Chromatography 1999, 847, 91-102, most of the prior art only deals with the analysis of few, specific components, for example acids or sugars. Most of the known processes limit a comprehensive biochemical analysis by the following facts: (a) they are not open, i.e. it is only possible to obtain information on metabolites which are already known, (b) they are very labor-intensive since they are frequently based on individual assays; (c) their resolution is only very poor, and the result is therefore a profile which only has a very low degree of complexity and contains little information, (d) they only cover metabolites in a cell in only one status, for example a growth phase or under the effect of a stress factor, and dynamic changes in the cell can therefore not be monitored, or (e) they only cover few of the classes of substances present in the organism, for example only sugars or only fatty acids or only certain molecular weights, but not the broad range of polar or unpolar substances. There exist only first attempts to demonstrate that metabolites not only constitute intermediate or end products, but also act as sensors and regulators.
The comprehensive and quantitative determination of the metabolites and constituents present in an organic sample (independently of whether this determination is limited to various classes of substances, given developmental stages or types of material, i.e. independently of whether it takes the form of metabolic fingerprinting, metabolic profiling or metabolomics) enables the direct study of immediate effects of growth, development or stress on the entire organism or parts thereof and is thus helpful as essential part of functional genome analysis in the determination of gene functions. Processes for analyzing metabolic profiles, in particular when they are suitable for analyzing substantial numbers of samples, permit the study of the complexity of the regulatory interactions at all levels and all stages and, last but not least, the assessment of the safety and value of genetically modified organisms.
The most advanced aspect is the determination of complex metabolic profiles in diagnostic screens, but first profiles have also recently been described for plants (for a review see Trethewey, Curr. Opin. Plant. Biol. 1999, 2, 83-85). Thus, Sauter (ACS Symposium Series 1991, 443 (Synth. Chem. Agrochem. 2), American Chemical Society, Washington, D.C., 288-299) demonstrates the modification of constituents in barley following treatment with various herbicides. Sauter describes the weighing of frozen plant samples and extraction thereof with 100% ethanol as advantageous method. Following filtration, drying and silylation, the samples can be separated via capillary columns. Between 100 and 200 signals were detected and identified with the aid of reference substances via their retention coefficients in gas chromatography (GC) or via gas chromatography/mass spectrometry analysis (GC/MS).
Fiehn, Nature Biotechnology 2000, 18, 1157-1161 describes the quantification of 326 substances in Arabidopsis thaliana leaf extracts. To compare four different genotypes, frozen plant samples were homogenized in a complicated procedure, extracted with 97% methanol and brief heating, and, after addition of chloroform and water, a multi-step procedure gave a polar and an unpolar phase which were then analyzed by LC/MS and GC/MS (see also Fiehn, Anal. Chem. 2000, 72, 3573-3580; http://www.mpimp-golm.mpg.de/fiehn/blatt-protokoll-e.html). Following a very similar method, Roessner, The Plant Journal 2000, 23, 131-142, extracts plant constituents with methanol and compares the profiles of polar metabolites of in-vitro potato plants and potato plants grown in soil.
Gilmour, Plant Physiology 2000, 124, 1854-1865 extracts sugar from lyophilized Arabidopsis leaves in 80% ethanol following incubation for 15 minutes at 80° C. and incubation overnight at 4° C. Strand, Plant Physiology 1999, 119, 1387-1397 extracts soluble sugars and starch twice in succession, likewise at 80° C. and for 30 minutes and in 80% ethanol with Hepes, pH 7.5. The material is then reextracted twice at this high temperature to improve the result of the extraction, once with 50% ethanol/Hepes, pH 7.5, and once with Hepes, pH 7.5.
These methods described in the prior art only permit limited automation which, moreover, can only be realized in the form of a complex procedure. In particular the processing of large sample numbers, the determination of the effect of a variety of stress factors on the metabolism of the organisms or the observation of dynamic processes, which requires a continuous analysis of samples during windows which are often very short, require processes    (a) which are rapid, i.e. for example that fixing and analysis of the samples is effected within a short period after sampling,    (b) which are highly reproducible, i.e. for example that an analysis carried out with a large number of different samples gives results within a very narrow error margin,    (c) which are simple to handle, i.e. for example that the process can be automated and does not require complex or laborious procedures,    (d) which are open, i.e. for example that a large number of substances can be analyzed, and/or    (e) which are sensitive, i.e. for example that the analysis identifies even small changes in substance concentrations and small amounts of substance.
Many processes have the disadvantage that they are only suitable for the analysis of small sample numbers. With a larger number of samples, it is not possible to ensure sample stability, and thus the reproducibility of the results. A comprehensive continuous analysis of biological material, for example animal samples or plant samples, or for example the interaction between a substance, or substances, and organisms in complex systems and their course over time is thus not possible.